LDPC decoder hard decision not producing expected result

Good day everyone, I'm currently running a project to investigate the performance of LDPC error correction technique in WSN. However, after I drew the blocks in simulink, the LDPC decoder is not producing expected result when i monitor the BER. I had troubleshoot this problem for a few weeks yet i still couldn't solve it. Can anyone please kindly help me? I'm really in a hot soup right now.

[Walter Roberson]

It would help to include your model file. It would also help to describe what you expect to see and what you see instead.

[Wen Pang]

Thank you for the reminder. As when i run the model with T=10, it provides me with error rate which is exactly equal to the number of bit that i run which means that LDPC is not correcting anything.

[Walter Roberson]

There must be something wrong with your model. The higher the error rate in your BSC, the higher the error rate on the uncorrected channel [error display further to the right] (which is as expected), but the lower the error rate after correction [error display further to the left]. Error rate with probability 0 in BSC is about 91% when it should come out as 0.

Is it possible this has to do with the number of samples of delay in the comparison?

[Wen Pang]

Thanks for the reply. That is the main problem that is bothering me right now. When I connect a simout to show the output of the corrected part, it was something totally different from the input. i think that is the main reason why it showed such high of error rate. I am running out of idea on how to correct it. Do you have any idea on how to solve it? Your effort is much appreciated.